It is well known that temperature is an important factor in the photosynthetic rate of oceanic phytoplankton communities (Li, Reference Li1985). The relationship between the photosynthetic rate and temperature differs with different phytoplankton sizes (Andersson et al., Reference Andersson, Haecky and Hagström1994; Shiomoto et al., Reference Shiomoto, Tadokoro, Monaka and Namba1997), though information about this relationship is very limited in the field. One study in the open ocean suggests that picoplankton (<2 µm) have a better ability of acclimation to fluctuations in temperature than larger phytoplankton (Shiomoto et al., Reference Shiomoto, Tadokoro, Monaka and Namba1997). In contrast, such results were not found in a coastal study (Andersson et al., Reference Andersson, Haecky and Hagström1994). The size composition of phytoplankton has much influence on the food chain in the marine ecosystem (e.g. Lalli & Parsons, Reference Lalli and Parsons1993). On the other hand, global warming is advancing (e.g. IPCC, 2001). It is thus necessary to accumulate information about the relationship between size-fractionated photosynthesis and temperature in forecasting future marine ecosystems.

This study was conducted in the western subarctic North Pacific during a cruise of the RV ‘Hokko Maru’ belonging to the Hokkaido National Fisheries Research Institute during the winter of 2006. Stations 1, 3 and 4 were located in the ocean and Station 2 was located 8 miles off the coast of Hokkaido, Japan (Table 1). Photosynthetic rates were measured at four different temperatures by the simulated in situ method using the ¹³C uptake technique (Hama et al., Reference Hama, Miyazaki, Ogawa, Iwakuma, Takahashi, Otsuki and Ichimura1983). Seawater samples were collected from the surface using a pumping up system attached to the ship. The samples (2 l) were dispensed into two 2-l polycarbonate bottles at each temperature. For seawater samples incubated at different temperatures from the in situ temperatures, the bottles were pre-incubated in incubators adjusted to the three different temperatures for about 1 hour under sunlight. The three temperatures were nearly as follows: 5, 10 and 15°C at Stations 1, 2 and 4, and 10, 15 and 20°C at Station 3. The seawater samples were then enriched with ¹³C-NaHCO₃ (99 atom% ¹³C; Shoko Co. Ltd.) to about 10% of the total inorganic carbon in ambient water and incubated for 6–8 hours under sunlight. For the seawater samples incubated at the in situ temperature, the bottles were held in a deck incubator for 6–8 hours maintained at a constant temperature by continuously flowing near-surface seawater, after inoculation with ¹³C-NaHCO₃ following pre-incubation for about 1 hour. Temperatures were monitored every 10 minutes during the incubations and nearly constant temperatures within ±2°C were confirmed in every case. The temperatures of the flowing seawater for incubation at the in situ temperature were about 2°C higher than the in situ temperatures.

Table 1. Environmental factors at the surface of the four stations.

^aOceanic station.

^bCoastal station.

The fractionation of samples into size-classes was performed after incubation. Immediately following incubation, 0.5-l seawater samples in bottles were directly filtered through pre-combusted (450°C for 4 hours) 47-mm Whatman GF/F filters (~0.7-μm pore size: total). The 0.5-l seawater remaining in the bottles after the above filtering was filtered through Nuclepore filters with pore sizes of 2 and 10 µm, and the filtrate was re-filtered through 47-mm Whatman GF/F filters (for the <2 or <10 µm size fraction). The particulate matter on the Whatman GF/F filters was rinsed with pre-filtered seawater and the filters were immediately frozen and preserved for isotope analysis ashore. They were treated with HCl fumes for 4 hours to remove inorganic carbon and completely dried in a vacuum desiccator. The isotopic ratios of ¹³C to ¹²C and particulate organic carbon were determined using a mass spectrometer (ANCA SL, CerCon). The total inorganic carbon in the water was measured with an infrared analyser (Shimadzu TOC 5000). The photosynthetic rate was calculated according to the equation described by Hama et al. (Reference Hama, Miyazaki, Ogawa, Iwakuma, Takahashi, Otsuki and Ichimura1983). The photosynthetic rate of the 2–10 and >10 µm size fractions was obtained from the differences between the <10 and <2 µm size fractions and between the total and <10 µm size fraction, respectively.

Surface temperature and salinity were measured with a thermometer and Guildline Model 8400A AUTOSAL, respectively. Nutrient concentrations were measured with a Bran & Luebbe Auto Analyser TRAACS 2000 after storage at –20°C. On-deck photon fluxes (PAR: photosynthetically active radiation) were monitored every two minutes with Alec model MDS MKV/L quantum sensor during the observation period. Seawater samples for measuring chlorophyll-a (Chl a) concentration were obtained from the same seawater samples for the measurement of photosynthetic rates. The procedure used for size-fractionation was the same as that used for photosynthesis. Chl a concentrations were determined with a Turner Designs 10-AU fluorometer according to Parsons et al. (Reference Parsons, Maita and Lalli1984) for samples extracted with N, N-dimethylformamide (Suzuki & Ishimaru, Reference Suzuki and Ishimaru1990). Calibration of a fluorometer was performed with a commercially prepared Chl a standard (Sigma Chemical Co.).

Salinity values were less than 34 psu at every station (Table 1), showing that every station was located in the subarctic region (Favorite et al., Reference Favorite, Dodimead and Nasu1976). Station 1 was located in the south-western part of the Western Subarctic Gyre and Station 4 to the south-west of Station 1. Station 2 with the lowest temperature and salinity was located in Coastal Oyashio Water (Hokkaido Fisheries Experimental Station, 2006) and Station 3 with the highest temperature and salinity was located in the warm core ring (Japan Coast Guard, 2006). Macronutrient concentrations were high at every station (Table 1). The PAR values (198–481 µmol of photons m⁻² s⁻¹) at all stations were higher than the values at which photosynthetic rates were saturated in the subarctic region (100–200 µmol of photons m⁻² s⁻¹; Andersson et al., Reference Andersson, Haecky and Hagström1994; Boyd et al., Reference Boyd, Whitney, Harrison and Wong1995). The photosynthetic rates obtained in this study were not considered to be limited by macronutrients and light. The short term effect of temperature increases in this study may not be relevant to longer time scales associated with temperature changes in the sea. However, photosynthetic rates in this study definitely show the photosynthetic potential of phytoplankton. Total Chl a concentrations were less than 0.6 µg l⁻¹ (Table 1). The shares of the smallest size (<2 µm) were highest at the three oceanic stations (52–65%), whereas the shares of the three sizes were nearly equal at the coastal station (30–35%) (Table 1).

Relationships between temperatures and Chl a-specific photosynthetic rates are shown in Figure 1. Photosynthetic rates of the <2 µm ranged from 0.5 to 2.4 µgC μgChl a ⁻¹ h⁻¹ at every station. The rates of the 2–10 µm ranged from 1.1 to 5.4 µgC μgChl a ⁻¹ h⁻¹ at Stations 1–3, and from 11 to 24 µgC μgChl a ⁻¹ h⁻¹ at Station 4. The rates of the >10 µm ranged from 0.1 to 5.0 µgC μgChl a ⁻¹ h⁻¹ at Stations 1, 2 and 4, and from 3.2 to 20 µgC μgChl a ⁻¹ h⁻¹ at Station 3. The remarkably high rates of the 2–10 and >10 µm sizes were caused by very low Chl a (<0.02 µg l⁻¹).

Fig. 1. Photosynthetic rates (μgC μgChl a ⁻¹ h⁻¹) of the <2, 2–10 and >10 µm sizes in relation to temperature (°C) deviation from minimum on board at Stations 1, 2, 3 and 4. The minimum incubated temperatures in the incubation experiments at Stations 1, 2, 3 and 4 are 2.2, 1.9, 1.6 and 2.1°C higher than the in situ temperatures, respectively. Solid lines represent changes in mean values.

Increases were observed in photosynthetic rates with a rising temperature for the <2 µm size at Stations 1, 2 and 3, and significantly linear relationships were found between the photosynthetic rates and temperatures (Station 1: r = 0.88, N = 8, P < 0.05; Station 2: r = 0.99, N = 8, P < 0.05; Station 3, r = 0.99, N = 8, P < 0.05; Figure 1). In contrast, at Station 4, saturation was found in the rates after 5.6°C higher than the minimum temperature. The rates of the 2–10 µm size increased with rising temperatures and nearly saturated at about 6°C higher than the minimum temperatures at Stations 1, 3 and 4 (Figure 1). In contrast, an almost linear increase was observed in the rates with rising temperatures at Station 2. A significant linear relationship was found between the rates and temperatures at this station (r = 0.97, N = 8, P < 0.05). For the >10 µm size, neither linear increases nor saturation followed by increases were observed in the relationships between the temperatures and the rates at Stations 1, 3 and 4 (Figure 1). At Station 2, a gentle increase was observed in the rates to 7.4°C higher than the minimum temperature and a remarkable increase was observed after 7.4°C higher than the minimum temperature.

The results obtained at Stations 1, 3 and 4 in the oceanic area show that the photosynthetic rates of middle-size phytoplankton saturated at lower temperatures than small size phytoplankton, meaning that small phytoplankton have better ability to acclimate to rapid fluctuations in temperature than middle-size phytoplankton. The results also show that large phytoplankton have inadequate ability to acclimate to rapid fluctuations in temperature. It is thus suggested that in the oceanic area the smaller phytoplankton are, the better their ability to acclimate to rapid fluctuations in temperature. Iron is considered to be a primary factor for phytoplankton, especially large phytoplankton, and the concentration is substantially lower in the ocean than along the coast (Martin et al., Reference Martin, Gordon, Fitzwater and Broenkow1989). The larger the phytoplankton are the more iron they require (Sarthou et al., Reference Sarthou, Timmermans, Blain and Tréguer2005). It is thus implied that large phytoplankton are readily susceptible to iron-limitation in the ocean, whereas small phytoplankton are not. Iron limitation will interfere with photosynthesis and hence acclimation to temperature fluctuation. It is therefore possible that, in the open ocean, the larger the phytoplankton are the more difficult it is for them to acclimate to temperature fluctuation. The results in this study indicate this possibility.

On the other hand, saturations were not observed in photosynthetic rates for every size of phytoplankton at Station 2 in the coastal area. Significant linear relationships were obtained between photosynthetic rates and temperatures for the <2 and 2–10 µm sizes (P < 0.05) and a rapid increase was observed after 7.4°C higher than the minimum temperature for the >10 µm size. In order to compare the effect of temperatures on photosynthetic rates of the three sizes, Q₁₀ was calculated by the following equation (Berry & Raison, Reference Berry, Raison, Lange, Nobel, Osmond and Ziegler1981):

(1)

where r = photosynthetic rate, t = temperature (°C). Mean photosynthetic rates at the minimum temperature and 13.3°C higher than the minimum temperature were used for the calculation. Q₁₀ values of up to 13.3°C for the <2 and 2–10 µm sizes were 2.3 and 1.3, respectively. For the >10 µm size, Q₁₀ value after 7.4°C higher than the minimum temperature was calculated to be 2.9 using mean rates at 7.4 and 13.3°C higher than the minimum temperature. No substantial difference was found between the values of the three sizes. It is thus possible that there is no significant difference in acclimating ability between sizes in coastal phytoplankton, though remarkable temperature rises are necessary for large phytoplankton to display their ability to the full.

Linear increases were observed in the photosynthetic rates up to 12°C for the <2 µm size at Stations 1 and 3 (Figure 1). Linear increases were also found in the rates up to 6°C for the <2 µm size at Station 4 and 2–10 µm size at Stations 1, 3 and 4. Q₁₀ values were calculated using mean photosynthetic rates at minimum and maximum temperatures for the linear increases. The values of the <2 and 2–10 µm sizes were 1.9–2.4 and 3.0–4.2, respectively. The values of the 2–10 µm size were somewhat higher than those of the <2 µm size. The values of the <2 µm size were almost equal to those of the same size in the coastal area (2.3). Q₁₀ values of the phytoplankton community converted from activation energy (Ea = 74 lnQ₁₀; Berry & Raison, Reference Berry, Raison, Lange, Nobel, Osmond and Ziegler1981) reported by Li (Reference Li1985) are 1.3–2.4 in the eastern subarctic North Atlantic in summer. The Q₁₀ values of the <2 µm size in this study are almost equal to Li's values. This coincidence is possibly due to the dominance of small phytoplankton in the oceanic phytoplankton community (e.g. Shiomoto et al., Reference Shiomoto, Tadokoro, Monaka and Namba1997). According to Andersson et al. (Reference Andersson, Haecky and Hagström1994), Q₁₀ values of the <2, 2–10 and >10 µm sizes were 7.1, 1.2 and 2.5 in the northern Baltic Sea in spring, respectively. The values of the 2–10 and >10 µm sizes were nearly equal to those obtained in this study and Li (Reference Li1985), whereas the value of the <2 µm size was substantially higher than those obtained in this study and Li.